All animal procedures were performed in compliance with the protocols approved by University of Michigan Institutional Animal Care and Use Committee in accordance with NIH OLAW policy and Guide for the Care and Use of Laboratory Animals, 8^th ed. Male wild-type C57BL/6J, Nox4 ^-/- (B6.129-Nox4^tm1Kkr /J), and Apoe ^-/- (B6.129P2-Apoe^tm1Unc /J) mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Littermate male mice were used in all experiments. Mice were housed in ventilated cages at 22°C with 12-hour light/dark cycle and free access to food and water. For atherosclerosis analysis Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice were fed standard rodent chow until 8-weeks-old, then fed Western type diet (TD.88137; Envigo, Madison, WI) for 12 weeks. For assessment of atherosclerosis in aging, mice were kept on standard rodent chow until 13-month-old, then fed Western diet for additional 12 weeks.

Monocytes were isolated from Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice as described (23). Briefly, mice were euthanized with inhaled isoflurane overdose, the spleens were dissected and mechanically dispersed. Cell suspension was passed through a 70 µm cell strainer (Corning, Corning, NY) and incubated in RBC lysis buffer (15.5 mM NH4Cl, 1 mM KHCO3, 10 µM EDTA, pH7.3). Cells were washed and cultured in RPMI1640 medium supplemented with 10% fetal bovine serum, 10 ng/mL M-CSF and antibiotic/antimycotic solution (Thermo Fisher; Waltham, MA) in a 5% CO₂ incubator at 37°C. Cells were treated with either 20 ng/mL IFNγ (PeproTech; Cranbury, NJ) and 10 ng/mL LPS (Sigma-Aldrich; St. Louis, MO), or 20 ng/mL IL4 (PeproTech) for 24 hours to induce M[IFNγ+LPS] (M1) or M[IL4] (M2) phenotypes, respectively. Cells treated with vehicle were used as M0 controls. Alternatively, cells were pretreated with 10 μM GKT137831 (Cayman Chemical; Ann Arbor, MI) for 30 minutes prior other treatments.

Mice were euthanized with inhaled isoflurane overdose and aortas were dissected, fixed and atherosclerotic lesion area determined as previously described (24). Longitudinally opened aortas were stained with 1% oil red O and 0.1% toluidine blue (Sigma). Digital images of stained aortas were analyzed with NIH ImageJ 1.54f (Bethesda, MD). Mouse hearts and proximal brachiocephalic artery were dissected, embedded in O.C.T. compound (Sakura Finetek; Torrance, CA), and snap-frozen in liquid nitrogen. Transverse 10 μm serial sections of aortic root and brachiocephalic artery were collected and processed as described before (24).

Picrosirius red staining was performed with brachiocephalic artery cryosections as previously described (25). Briefly, sections were fixed in 3.7% paraformaldehyde stained using Picrosirius Red Stain Kit (Abcam; Waltham, MA). Images were acquired with Revolve R4 microscope (Echo; San Diego, CA) in brightfield mode and analyzed with NIH ImageJ software.

Immunofluorescence staining was performed as previously described (24, 26). Cryosections were fixed in acetone, permeabilized in 0.1% Triton X-100 and immunostaining was done using antibodies against 8-OHdG, NOX4, CCL2, IL1b, TOM20-AlexaFluor488 (Abcam), CD68 (Thermo Fisher); ACTA2-FITC (Sigma); CD11b (Abnova; Walnut, CA); IL6 (Cell Signaling Technology; Danvers, MA); ATP5G-AlexaFluor488, CD80-AlexaFluor488, CD68-Cy3, CD163-AlexaFluor594, CD206-AlexaFluor488 (Bioss). The goat anti-rabbit antibodies conjugated to AlexaFluor488 or AlexaFluor594, or rabbit anti-goat AlexaFluor594 secondary antibodies were used when appropriate. Sections were mounted with ProLong Gold reagent with DAPI (Thermo Fisher), fluorescence images were taken with Nikon Microphot-FX microscope and analyzed using NIH ImageJ.

Brachiocephalic artery cryosections were analyzed immediately after collection. Cellular ROS levels were determined with dihydroethidium (Thermo Fisher) fluorescence and mitochondrial ROS with MitoSOX Red (Thermo Fisher) fluorescence as previously described (12, 14). Fluorescence images were taken with Nikon Microphot-FX microscope using 510 nm excitation/580 nm emission filters. Images were analyzed using NIH ImageJ software. Control sections incubated with PEG-SOD were used for background/autofluorescence correction.

AmplexRed assay (Thermo Fisher) was used to determine hydrogen peroxide release from freshly dissected aorta samples as previously described (27). Briefly, dissected aortas were gently minced and incubated in AmplexRed working solution for 30 minutes. Fluorescence was measured in sample aliquots using Spectramax iD5 Multi-Mode Microplate Reader (Molecular Devices; San Jose, CA) and the amount of H₂O₂ was determined using standard curve. Dry tissue weight was determined and used to normalize H₂O₂ levels.

Superoxide levels in cultured macrophages were determined with HPLC measurement of 2-hydroxyethidium levels as described previously (12) based on the method described by Zielonka et al. (28). Briefly, Samples were analyzed using Agilent 1100 HPLC system equipped with Kinetex 2.6 µm C18 100Å 100x4.6 mm LC column (Phenomenex; Torrance, CA). The amount of 2-hydroxyethidium was determined using oxyethidium (Noxygen Science Transfer & Diagnostics GmbH; Elzach, Germany) standard curve and normalized to the sample protein concentration measured with Pierce BCA Protein Assay Kit (Thermo Fisher).

Mitochondrial ROS in cultured macrophages were determined with MitoSOX Red (Thermo Fisher) fluorescence as described (14, 25). Briefly, cells were grown on tissue culture treated glass chamber slides (Corning Life Sciences; Tewksbury, MA), then, after treatment, cells were incubated with 10 µM MitoSOX Red at 37C° for 30 minutes and confocal images taken with Echo Revolve R4 microscope. Single cell images were analyzed with NIH ImageJ software. Control cells incubated with PEG-SOD were used for background/autofluorescence correction.

Analysis of peripheral blood mononuclear cells was performed as described previously (29). Blood samples were collected in Lithium Heparin BD Vacutainers (BD Biosciences; Franklin Lakes, NJ). Red blood cells were lysed with RBC Lysis Buffer (Thermo Fisher). The remaining cells were passed through a 70 µm cell strainer (Corning) and washed with PBS. Cells were blocked with rat anti-mouse CD16/CD32 antibody mix (BD Biosciences), then stained with Zombie UV Viability Dye (BioLegend; San Diego, CA). CD11b-APC, LY6C-FITC (Miltenyi Biotech; Gaithersburg, MD), and CD115-AlexaFluor532 (Novus Biological; Centennial, CO) antibodies were used for staining. The samples were processed with MoFlo Astrios EQ Cell Sorter (Beckman Coulter; Brea, CA) and analyzed using FlowJo v10.8 (BD Biosciences). Peripheral blood mononuclear cells were identified as CD11b⁺CD115⁺LY6C⁺ fraction of all cells. Based on the LY6C fluorescence intensity LY6C^hi and LY6C^lo cells were identified ( Supplementary Figure 1 ).

Analysis of aortic atherosclerotic lesions-derived single-cell suspension was done as reported previously (24). Briefly, dissected aortas were gently minced and dissociated in collagenase type I (400 U/mL), collagenase type XI (120 U/mL), hyaluronidase (60 U/mL), and DNase I (60 U/mL) solution. Cell suspension was passed through 70 µm cell strainer (Corning) and washed with FACS buffer. A negative control cell pull was left unstained. Cells were blocked with rat anti-mouse CD16/CD32 antibody mix (BD Biosciences), then stained with Ghost Dye Red 710 Viability Dye (Cell Signaling Technology). The staining mix containing CD11b-PE, F4/80-APC-Cy7, CD86-PerCP-Cy5.5, CD206-PE-Cy7 (BioLegend), CD38-FITC, and EGR2-APC (Thermo Fisher) antibodies was applied to the cells for 1 hour. Samples were washed and resuspended in FACS buffer. Flow cytometry analysis was performed with Aurora Spectral Analyzer (Cytek; Freemont, CA). UltraComp eBeads Compensation Beads (Thermo Fisher) bound with corresponding antibodies were used as single-stained controls. An unmixing matrix with autofluorescence extraction was calculated using SpectroFlo software (Cytek) and the samples were run using the live unmixing. Unmixed fcs files were analyzed using FlowJo v10.8 (BD Biosciences). Gating strategy shown in the Supplementary Figure 2 .

Cultured spleen-derived monocytes, treated to induce M0, M[IFNγ+LPS] or M[IL4] phenotype, were lifted from the culture plates using Dispase (Thermo Fisher), washed with PBS, blocked with rat anti-mouse CD16/CD32 antibody mix (BD Biosciences), and stained with Ghost Dye Red 710 Viability Dye (Cell Signaling Technology). Cells were incubated in the antibody mix containing CD11b-PE, F4/80-APC-Cy7 (BioLegend), CD38-FITC, and EGR2-APC (Thermo Fisher) antibodies for 1 hour. Flow cytometry analysis was performed with Aurora Spectral Analyzer (Cytek) as described above.

Macrophage mitochondrial bioenergetics was assessed using Seahorse XFe96 analyzer (Agilent Technologies; Santa Clara, CA) as described previously, with modifications (14). Live macrophage numbers were determined with Countess 3 FL Automated Cell Counter and 70000 cells/well were plated in Seahorse XFe96 cell culture microplates (Agilent). Attached cells were treated with either vehicle, 20 ng/mL IFNγ (PeproTech) and 10 ng/mL LPS (Sigma-Aldrich), or 20 ng/mL IL4 (PeproTech) for 24 hours. Before the assay, cells were washed, and media replaced with Seahorse XF RPMI medium supplemented with 10 mM glucose and 2 mM glutamine. Culture plate was placed in the CO₂-free incubator for 1 hour before the test. Seahorse XFe96 Extracellular Flux Cartridge was hydrated with sterile water overnight in CO₂-free incubator. Water was replaced with XF Calibrator 1 hour before the test. The oxygen consumption (OCR) and extracellular acidification rates (ECAR) were determined using Seahorse XFp Cell Mito Stress Test Kit (Agilent) consisting of 3 consecutive injections: 2 µM oligomycin (Oligo), 2 µM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP), and 1 µM rotenone with antimycin A (Rot/AA). Immediately after the test, cells were washed, lysed with M-PER Protein Extraction Reagent (Thermo Fisher), and protein concentration measured with Pierce BCA Protein Assay Kit (Thermo Fisher). The flux rates were normalized to µg protein per well. Mitochondria bioenergetics parameters were calculated as described previously (14). Metabolic profiles, mitochondrial and glycolytic contributions to ATP production rate were determined as described in (30).

Plasma cholesterol and triglycerides levels concentration were determined with AMS Liasys 330 Chemistry Analyzer (AMS Alliance; Weston, FL).

IL1β levels in conditional media were determined with Mouse Interleukin-1beta ELISA Kit (Thermo Fisher) following the manufacturer’s protocol.

Total RNA was extracted from cultured macrophages using RNeasy Micro Kit (Qiagen; Germantown, MD), then cDNA was synthetized using iScript cDNA Synthesis Kit (Bio-Rad; Hercules, CA). Real-time PCR was performed with TaqMan GEN Expression Assays for Nox4 (Mm00479246-m1) and 18S (Hs99999901_s1), Universal PCR master mix (Thermo Fisher) and analyzed using 7500 Fast Real-Time PCR System (Thermo Fisher). Target RNA relative expression level was calculated by normalization to 18S rRNA expression level.

All analyses were done using Prism 9 (GraphPad Software; La Jolla, CA). All data were tested for normality with D’Agostino & Pearson test. Data were analyzed using unpaired t-test or one-way ANOVA with the Tukey multiple comparisons test, when appropriate. The results of OCR and ECAR measurements were analyzed using a two-way repeated measures ANOVA. All data presented as mean ± SEM. Differences were considered significant at p<0.05.

The expression of NOX4 is increased in mouse and human atherosclerotic lesions with aging (14). To examine the impact of NOX4 on aging-associated atherosclerosis we initially determined atherosclerosis burden in four groups: young (5-month-old) and aged (16-month-old) Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice. All the mice were fed Western diet for 3 months. The aortic atherosclerotic lesion area was comparable between young Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice. Aged mice of both genotypes had significantly increased lesion area with Apoe ^-/- showing 195% increase and Nox4 ^-/-/Apoe ^-/- mice showing 82% increase. However, the lesion area was significantly lower in Nox4 ^-/-/Apoe ^-/- compared with Apoe ^-/- mice (p<0.0001; Figure 1A ). Consistent with that, analysis of atherosclerotic lesion volume integrated from aortic sinus serial sections indicated similar lesion volume in young mice and significantly increased lesions in aged Apoe ^-/- but not in aged Nox4 ^-/-/Apoe ^-/- mice (p<0.0001 and p=0.1372, respectively; Figure 1B ). Plasma cholesterol or triglyceride levels were not significantly different between young or aged Apoe ^-/- mice and their Nox4 ^-/-/Apoe ^-/- counterparts ( Figures 1C, D ). Mouse body weights increased with age but were not different between genotypes ( Supplementary Figure 3A ). There were no significant differences in plasma glucose or liver enzymes ( Supplementary Figures 3B–D ). Therefore, the observed increase in atherosclerosis cannot be attributed to metabolic changes, which underscores the role of NOX4 in the age-related progression of atherosclerotic lesions.

The progression of atherosclerotic lesions in the brachiocephalic (innominate) artery of older Apoe ^-/- mice is analogous to advanced human atherosclerotic disease (31). For that reason, we also evaluated atherosclerotic lesions in brachiocephalic/common carotid arteries of young and aged Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice. We found no significant differences in the cross-sectional area of atherosclerotic lesions in the brachiocephalic arteries of young mice, consistent with our findings in aortic sinus lesions. However, as the Apoe^-/- mice aged, the lesion area increased significantly, while the Nox4^-/-/Apoe^-/- mice showed no significant increase (p=0.0432 and p=0.8422, respectively; Figure 1E ). Additionally, Apoe^-/- mice showed a disintegrated fibrous cap and significantly lower plaque collagen content, as revealed by Picrosirius red staining ( Figure 1F ). In contrast, Nox4-deficient mice had preserved lesion fibrous cap and higher plaque collagen content compared with Apoe ^-/- mice. Although aging was associated with a slight decrease in collagen content, this change did not reach statistical significance ( Figure 1F ). These results suggest that aging is associated with a significant increase in atherosclerosis burden and the development of advanced plaque in Apoe ^-/- mice. In contrast, Nox4 deletion reduced lesion size and preserved plaque integrity, indicating a critical role for NOX4 in aging-associated atherosclerosis.

Aging is linked to increased cellular and mitochondrial ROS, and oxidative DNA damage within atherosclerotic lesions of Apoe ^-/- mice (12, 14). Consistent with this, the ROS levels, determined by DHE fluorescence in brachiocephalic artery atherosclerotic lesions from Apoe ^-/- mice, significantly increased with age (p=0.0003; Figure 2A ). Young Nox4 ^-/-/Apoe ^-/- exhibited ROS levels similar to those in Apoe ^-/- mice. However, in aged Nox4 ^-/-/Apoe ^-/- mice ROS levels were significantly lower than in their aged Apoe ^-/- counterparts (p=0.0044).

Similarly, mitochondrial ROS levels in brachiocephalic artery atherosclerotic lesions were not different between young Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice when measured by MitoSOX fluorescence but increased significantly in aged Apoe ^-/- mice (p=0.0013; Figure 2B ). Conversely, aged Nox4 ^-/-/Apoe ^-/- mice did not exhibit a significant change in mitochondrial ROS levels compared with young mice (p=0.5271; Figure 2B ).

Given that NOX4 activity primarily generates hydrogen peroxide (32), we measured H₂O₂ release in freshly isolated aortas using the AmplexRed assay. Although no significant differences were observed in H₂O₂ levels in the aortas from young mice, hydrogen peroxide levels increased significantly in aged Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice (p<0.0001 and p=0.0218, respectively; Figure 2C ). Importantly, H₂O₂ levels in aged Nox4 ^-/-/Apoe ^-/- aortas were significantly lower than in aged Apoe ^-/- mice (p<0.0001). These results suggest that NOX4 is a key contributor to ROS production in aging-associated atherosclerosis.

The augmented oxidative stress in aging-associated atherosclerosis causes oxidative damage to both nuclear and mitochondrial DNA in vascular smooth muscle cells (VSMC) (12, 14). Assessment of 8-hydroxy-2'-deoxyguanosine (8-OHdG) levels in brachiocephalic artery atherosclerotic lesions showed a significant increase in 8-OHdG immunofluorescence colocalization with nuclear and mitochondrial DNA in aged Apoe ^-/- compared with young mice ( Figures 2D, E ). However, the age-associated increase in nuclear and mitochondrial 8-OHdG levels did not occur in aged Nox4^-/-/Apoe^-/- mice, supporting the notion that high NOX4 levels increase mitochondrial oxidative stress and cause DNA damage in aging atherosclerosis.

Previous studies showed that inhibition of NOX4 or mitochondrial ROS in VSMC during aging atherosclerosis reduces vascular inflammation (14, 15). To further investigate this, we quantified aging-induced NOX4 expression in atherosclerotic plaque macrophages. Our analysis of brachiocephalic artery sections in aged Apoe ^-/- revealed a significant increase in expression of immunoreactive NOX4 colocalized with CD68⁺ cells compared with their young counterparts (a 230% increase; Figure 3A ). This increase was most prominent in macrophages associated with the fibrous cap that is consistent with the vulnerable plaque phenotype. Furthermore, the relationship between NOX4 and mitochondrial oxidative stress in aging atherosclerosis was supported by substantially increased colocalization of immunoreactive NOX4 with mitochondrial marker ATP5G in brachiocephalic artery sections from aged Apoe ^-/- mice (a 3.3-fold increase over young Apoe ^-/- mice; Figure 3B ).

CCL2-dependent monocytes recruitment to the atherosclerotic plaque is prominently involved in atherosclerosis progression and development of complications (33). Immunoreactive CCL2 expression was significantly higher in the atherosclerotic lesion sections from aged Apoe ^-/- compared with young Apoe ^-/- mice. The expression of CCL2 in aged Apoe ^-/- lesions was highest in fibrous cap-associated cells and medial SMC. In contrast, CCL2 expression in the plaques of young Nox4-deficient Apoe ^-/- mice was similar to that of young Apoe ^-/- mice and did not increase with age ( Figure 3C ). Flow cytometry analysis of peripheral blood cells showed that aged Apoe^-/- mice had higher percentages of LY6C^hiCD11b⁺CD115⁺ monocytes compared with young ones (p=0.0023; Figures 3D, E ). Contrary to that, the percent of LY6C^hi monocytes was not different in aged compared with young Nox4 ^-/-/Apoe ^-/- mice but was significantly lower than that in aged Apoe ^-/- mice (p=0.0033; Figures 3D, E ). Notably, the percentage of LY6C^lo monocytes was not different between young and aged mice ( Figure 3F ).

Consistent with that, immunofluorescent staining of brachiocephalic artery sections revealed that aged Apoe^-/- mice had significantly more lesion infiltrating CD11b macrophages than young mice or aged Nox4 ^-/-/Apoe ^-/- mice ( Figure 3G ). These findings suggest that increased NOX4- and mitochondria-derived ROS in aging facilitate monocyte mobilization and recruitment in atherosclerotic lesions. Immunofluorescent staining further revealed significantly higher expression of immunoreactive IL1β and IL6 in atherosclerotic lesion in aged Apoe ^-/- compared with young mice or aged Nox4 ^-/-/Apoe ^-/- mice ( Figures 3H, I ). Notably, higher expression of IL1β in aged Apoe ^-/- mice plaques was localized with the cells proximal to the fibrous cap and medial cells, whereas IL6 expression was higher in the plaque core cells and medial SMC. These observations indicate that increased mitochondrial oxidative stress in aging-associated atherosclerosis mediates monocyte mobilization/recruitment and intraplaque inflammation. Conversely, the deletion of Nox4 reduces plaque macrophage infiltration and inflammation.

To better understand the inflammatory phenotype of the atherosclerotic lesion macrophages, we conducted flow cytometry analysis of cells obtained from aortas of young and aged Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice fed Western diet (24). First, we determined the percentage of CD11b⁺F4/80⁺ macrophages in aortic cell suspensions. Consistent with the atherosclerotic lesion size, the proportion of macrophages was similar in young Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice ( Figure 4A ). However, the macrophage fraction of the aortic cells increased significantly with age and was 183% higher in aged Apoe ^-/- than in Nox4 ^-/-/Apoe ^-/- mice (p=0.0007; Figure 4A ).

In an effort to distinguish between classically (M[IFNγ+LPS]) and alternatively activated (M[IL4]) macrophages, we used specific markers for flow cytometry analysis - CD38 and EGR2, respectively (34). The proportion of CD38⁺ among CD11b⁺F4/80⁺ aortic cells was not different between genotypes and significantly increased with age in both Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice ( Figure 4B ). Conversely, the proportion of EGR2⁺ M[IL4] aorta macrophages was not different in young mice and significantly increased with age in Nox4 ^-/-/Apoe ^-/- but not in Apoe ^-/- mice ( Figure 4C ). This pattern was mirrored in the analysis of the CD86⁺ M[IFNγ+LPS] macrophages that demonstrated significant aging-associated increases in both genotypes, whereas the CD206⁺ M[IL4] aorta macrophage proportion significantly increased with age only in Nox4 ^-/-/Apoe ^-/- but not in Apoe ^-/- mice ( Figures 4D, E ). Moreover, the EGR2⁺/CD86⁺ macrophage ratio in atherosclerotic aortas was not different in young mice and increased with age solely in Nox4 ^-/-/Apoe ^-/- mice ( Figure 4F ). The relative increase of classically activated proinflammatory macrophages fraction was evident in atherosclerotic lesions of both genotypes. However, because the number of CD11b⁺F4/80⁺ macrophages in atherosclerotic aortas was significantly decreased in aged Nox4 ^-/-/Apoe ^-/- mice, the proportion of CD38⁺CD86⁺ macrophages was also relatively lower compared with that in aged Apoe ^-/- mice.

To corroborate these findings, we performed immunofluorescent staining of the brachiocephalic artery sections from young and aged Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice to determine the relative abondance of M[IFNγ+LPS] or M[IL4] macrophages within the atherosclerotic plaques. Immunofluorescence analysis revealed that young mice in both groups had a similar number of CD80⁺CD68⁺ (M[IFNγ+LPS]) cells ( Figure 4G ). However, the atherosclerotic plaques of aged Apoe ^-/- mice had a significantly higher proportion of CD80⁺CD68⁺ plaque cells compared with young mice or aged Nox4 ^-/-/Apoe ^-/- mice (p=0.0144). Furthermore, immunofluorescence staining for CD206 and CD163 displayed no differences in atherosclerotic plaques from young mice. Nonetheless, there was a substantially higher proportion of CD206⁺CD163⁺ cells in the brachiocephalic artery plaques from aged Nox4 ^-/-/Apoe ^-/- compared with aged Apoe ^-/- mice (p=0.0006; Figure 4H ). Taken together, these results suggest that the number of atherosclerotic lesion macrophages increases in aged Apoe ^-/- mice, predominantly comprising of classically activated pro-inflammatory macrophages. In contrast, Nox4 deficiency in aged Apoe ^-/- mice results in reduced macrophage infiltration in the plaque and increased proportion of pro-resolving macrophages. This combination leads to attenuated atherosclerotic lesion size and lower intraplaque inflammation.

Mitochondrial dysfunction and metabolic switch are implicated in pro-inflammatory priming of peripheral monocytes and subsequent changes in macrophage phenotypes within atherosclerotic plaques (8). The spleen is a source of monocytes that infiltrate atherosclerotic lesions (35), so we used spleen-derived monocytes from young and aged Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice to assess macrophage mitochondrial bioenergetics. We measured the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) to assess mitochondrial respiration and cellular metabolic rate in M0, M[IFNγ+LPS] or M[IL4] cells. Our results showed that mitochondrial respiration in control M0 and M[IL4] monocyte-derived macrophages from young mice was similar ( Supplementary Figures 4A, B, E, F ). However, with aging, there was a significant decline in OCR in both M0 and M[IL4] macrophages, and an increase in ECAR in M0 macrophages. Both basal and ATP-linked respiration were significantly lower in aged Apoe ^-/- M0 and M[IL4] as compared with the cells from young mice or aged Nox4 ^-/-/Apoe ^-/- mice ( Figures 5A, C ). Furthermore, mitochondrial respiration was lower in M[IFNγ+LPS] as compared with M0 or M[IL4] macrophages ( Supplementary Figure 4C ). In M[IFNγ+LPS] monocyte-derived macrophages, OCR and ECAR were similar in young Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice. However, aging significantly downregulated OCR and upregulated ECAR in Apoe ^-/- but not in Nox4 ^-/-/Apoe ^-/- monocyte-derived macrophages ( Supplementary Figures 4C, D ). Basal, maximal, and reserve respiration rates in aged Apoe ^-/- M[IFNγ+LPS] were significantly lower than that in aged Nox4 ^-/-/Apoe ^-/- macrophages ( Figure 5B ).

The ratio of OCR/ECAR greater than 4 indicates the preference for oxidative phosphorylation over glycolysis for energy production. In macrophages derived from aged Apoe ^-/- monocytes the OCR/ECAR ratio was below 4, which shows a preference for glycolysis, in contrast with cells from their young counterparts or Nox4-deficient mice ( Supplementary Figure 4G ). In line with that, metabolic profiling of M0, M[IFNγ+LPS] or M[IL4] macrophages derived from aged Apoe ^-/- monocytes exhibited a similar pattern characterized by lower OCR, higher ECAR, and a greater reliance on glycolysis ( Figures 5D–F ).

Our calculations of mitochondrial oxidative phosphorylation and glycolysis contribution to the ATP production also showed that aging significantly increased glycolytic ATP and reciprocally decreased mitochondrial ATP production in M0 and M[IFNγ+LPS] macrophages derived from aged Apoe ^-/- monocytes ( Figures 5C–H ). Taken together, these data suggest that higher NOX4 activity and mitochondrial oxidative stress in aging were synergistically associated with M[IFNγ+LPS] pro-inflammatory macrophage phenotype, reduced mitochondrial respiration and increased glycolysis. On the other hand, the absence of Nox4 in aging was associated with M[IL4] pro-resolving macrophage phenotype and preserved mitochondrial function.

In our previous studies we found that aging is associated with a significant increase in NOX4-dependent cellular and mitochondrial ROS in atherosclerotic lesions (13–15). In situ ROS measurements suggested that macrophages are significant contributors to the intraplaque oxidative stress and inflammation. To understand the role of oxidative stress in macrophage phenotype changes, we measured ROS levels in spleen-derived M0, M[IFNγ+LPS] or M[IL4] cells from young and aged Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice.

We first measured superoxide (O₂ ^•−) generation in macrophages by HPLC analysis of 2-hydroxyethidium. The O₂ ^•−-dependent levels of 2-OH-ethidium were not different in M0 or M[IL4] macrophages from young mice of both genotypes but were significantly higher in Apoe ^-/- than in Nox4 ^-/-/Apoe ^-/- M[IFNγ+LPS] macrophages ( Figure 6A ). The levels of O₂ ^•− in Apoe^-/– derived macrophages increased significantly with age. Similarly, O₂ ^•− levels associated with aging were significantly higher in Nox4-deficient M[IFNγ+LPS] macrophages. However, O₂ ^•− levels were lower in Nox4-deficient M0 and M[IL4] macrophages as compared with the aged Apoe^-/– derived macrophages ( Figure 6A ).

Immunofluorescence analysis of Apoe ^-/- macrophages indicated that mitochondrial NOX4 expression levels (NOX4 colocalized with TOM20) were significantly higher in aged than in young cells ( Figure 6B ). Congruently, Nox4 mRNA expression increased with aging in Apoe ^-/–derived monocytes by 80-fold on average ( Figure 6C ). However, Nox4 expression levels were not directly affected by macrophage polarization and did not differ between M0, M[IFNγ+LPS] or M[IL4] macrophages from either young or aged mice.

Analysis of MitoSOX fluorescence indicated no significant differences in mitochondrial ROS levels in macrophages from young mice ( Figure 6D ). M0, M[IFNγ+LPS] and M[IL4] macrophages from aged Apoe ^-/- mice had significantly higher mitochondrial ROS levels than cells from the young mice. In contrast, mitochondrial ROS increased with age only in M[IFNγ+LPS] Nox4 ^-/-/Apoe ^-/- macrophages but remained significantly lower than in aged Apoe ^-/- ( Figure 6D ). Similarly, M[IL4] aged Nox4 ^-/-/Apoe ^-/- macrophages had lower mitochondrial ROS levels than in aged M[IFNγ+LPS] Nox4 ^-/-/Apoe ^-/- macrophages. ( Figure 6D ). However, M[IL4] macrophages showed lower mitochondrial ROS levels than M[IFNγ+LPS] macrophages in both young and aged mice.

We conducted further tests to determine whether macrophage polarization is affected by NOX4-derived mitochondrial ROS. Flow cytometry analysis of spleen-derived monocytes from young and aged Apoe ^-/- and Nox4 ^-/-/Apoe ^-/- mice did not show differences in the percentage of CD38⁺ or EGR2⁺ cells in CD11b⁺ M0 macrophages ( Figures 6E, F ). The percentage of CD38⁺CD11b⁺ cells increased with age in mice of both genotypes but was significantly more so in Apoe ^-/- then in Nox4 ^-/-/Apoe ^-/- mice ( Figure 6E ). In contrast, the percentage of EGR2⁺CD11b⁺ M[IL4] was comparable between the genotypes in macrophages from young mice, but significantly increased in macrophages from aged mice. The increase was significantly higher in aged Nox4 ^-/-/Apoe ^-/- compared with Apoe ^-/- counterparts ( Figure 6F ).

Because increased mitochondrial ROS and dysfunction in aging leads to inflammasome activation (36) and is associated with pro-inflammatory macrophage phenotype (37), we assessed inflammasome activity by measuring IL1β levels in conditional media from M0, M[IFNγ+LPS], and M[IL4] monocyte-derived macrophages. There were no significant differences in IL1β levels in M0 or M[IL4] macrophages derived from monocytes of young or aged mice of both genotypes ( Figures 6G, I ). However, M[IFNγ+LPS] macrophages from aged Apoe ^-/- mice had significantly higher secreted IL1β levels than in macrophages from young Apoe ^-/- or aged Nox4 ^-/-/Apoe ^-/- mice ( Figure 6H ).

Our results indicate that increased NOX4 activity in aging induces mitochondrial oxidative stress and dysfunction, thereby inducing glycolytic metabolic switch, inflammasome activation and priming monocytes to acquire pro-inflammatory phenotype. However, if NOX4 activity is reduced during aging, mitochondrial function is preserved, inflammasome activation is reduced, and a higher percentage of monocytes acquire a pro-resolving phenotype.

We have previously demonstrated that using a pharmacological inhibitor that targets NOX4/NOX1 activity in aged Apoe ^-/- mice led to a decrease in atherosclerotic lesion size, VSMC mitochondrial ROS levels, and inflammation (14, 15). These studies underscored the role of VSMC NOX4 in the development of cardiovascular pathology associated with aging. To investigate the effects of NOX4 inhibitor on macrophage mitochondrial function and metabolic phenotype in atherosclerosis, we first measured the levels of mitochondrial ROS in macrophages derived from aged Apoe ^-/- mice ( Figure 7A ). The MitoSOX fluorescence was significantly higher in M[IFNγ+LPS] compared with M0 or M[IL4] macrophages. Treatment with GKT137831 significantly abrogated MitoSOX fluorescence in M[IFNγ+LPS] macrophages from aged Apoe ^-/- mice (p<0.0001; Figure 7B ). Consistent with increased ROS levels, mitochondrial OCR was substantially reduced in M[IFNγ+LPS] macrophages from aged Apoe ^-/- mice as compared with M0 or M[IL4] ( Figure 7C ; Supplementary Figures 5A, C ). However, GKT137831 treatment significantly improved mitochondrial function by increasing both OCR and maximal respiration ( Figures 7C, D ). Similarly, the GKT137831 significantly enhanced reserved respiration in M0 or M[IL4] and maximal respiration in M0 macrophages from aged Apoe ^-/- mice ( Supplementary Figures 5B, D ). Consistent with our observations in aged Nox4 ^-/-/Apoe ^-/- mice, the treatment with NOX4 inhibitor in aged Apoe ^-/- mice significantly reduced the number of CD68⁺CD80⁺ ( Figure 7E ) and increased the number of CD206⁺CD163⁺ ( Figure 7F ) macrophages observed in atherosclerotic lesion sections.

Taken together, these data suggest that similar to genetic deletion of Nox4, the inhibition of NOX4 activity reduces mitochondrial ROS levels and preserves function in M[IFNγ+LPS] macrophages, thereby limiting the expansion of pro-inflammatory macrophage population in the atherosclerotic lesion. Conversely, NOX4 inhibition significantly increases the proportion of pro-resolving macrophages, a trend that aligns with the reduction in vascular inflammation and atherosclerotic lesion area in aged Apoe ^-/- mice.